Statistical process control charts track coating processes by plotting data over time. Cpk is a statistical index that indicates whether a process will consistently produce coated units with a film having a thickness that is centered within tolerance limits. This invention relates to a novel computer-implemented process that anticipates the effect of Cpk for optimizing coatings usage and costs by reducing variability in film build thickness. The process minimizes the environmental impact of volatile organic compounds from the overspray of finishes being applied to painted parts. Improved quality is achieved by analyzing obvious non-random patterns of film variability. The system uses software that refers to the Industry and/or World Class Standards for Cpkxe2x80x2s. Optimum film build range and average values selected from the process and/or the coating applicators"" manufacturer guidelines are used for calculating the impact of changing coating thickness on all finished surfaces in terms of paint usage and costs.
Manufacturers of painted products, such as automotive bodies and/or furniture companies, assess paint film build thickness by various measurement tools. These tools include the Elcometer, the Pelt Gage or a Wet Gage. These devices measure film build thickness at specific points on the painted unit. Data from these measurements are then downloaded into a commercially available software database. Numerous statistical process control (SPC) and trend charts are generated from this data. These control charts include {overscore (X)} (average thickness vs. time); R charts (range of thickness vs. time) [APPENDIX A], and Individual Moving Range Charts [APPENDIX B]. Appendix A charts the film thickness average readings of a paint coating, taken Oct. 10, 2000 to Mar. 21, 2001, in millimeters. Appendix B charts the film build average values on automotive bodies taken Oct. 10, 2000 to Dec. 21, 2000.
Control limits are defined as a line (or lines) on a chart used for evaluating the stability of a process.
Typical control limits are plus or minus three standard deviation limits using at least 20 data points. When a point falls outside these limits, the process is said to be out of control.
Additionally, schematics of the painted unit [APPENDIX C] can be prepared which highlight or animate the painted surface areas"" compliance with either material and/or engineering coating specifications.
Process engineers review these charts and make corrective changes to the automation equipment, and any manual application equipment applying the coating. Considerable expenditure is spent on paint automation equipment and manual spraying techniques to ensure that the highest quality finish is produced at the lowest possible cost. This cycle repeats itself daily in many coating industries. Engineering reviews SPC charts and/or trend charts, and then adjusts automation based on historical data.
A statistically controlled condition exists when all special causes of film thickness variations have been eliminated with only common causes remaining. A xe2x80x9ccommonxe2x80x9d cause is a source of variation that affects all the individual thickness values of the process. An SPC chart that is described as xe2x80x9cin statistical controlxe2x80x9d possesses data values that neither surpass the control limits of the charts, nor possess non-random patterns or trends within the control limits.
Engineers refer to Process Capability indices (Cp and Cpk) generated by the SPC charts to evaluate the total range of a process""s inherent variation. Cp is defined as a measurement of the allowable tolerance spread divided by the actual 6"sgr" spread data. Cpk has a similar ratio to Cp but considers the shift of the mean film thickness relative to the central specification target.
Industry groups have set different control targets for what they consider capable processes. Raw data in software statistical control charts generate Cp and Cpk data. However, one manufacturer may consider the process in control when a Cpk of 1.33 is obtained, while another may seek a Cpk of 1.5. The Coating Applications Industry does not have a way to quickly identify the optimum statistical data from process data that will produce the given industry standard for a given Cpk value. Control Charts currently are used only for tracking purposes rather than control purposes.
Previous attempts at controlling processes within the Coatings Application Industry with only SPC (Statistical Process Control) charts suffer from a number of disadvantages:
(a) Engineers cannot calculate the minimal material usage for each painted surface area by increasing the process Cpkxe2x80x2s using commercially available SPC charts.
(b) Engineers cannot calculate the cost savings that can be realized for each painted surface area by increasing the process Cpkxe2x80x2s using commercially available SPC charts.
(c) Engineers only use SPC Control Charts for tracking rather than controlling purposes.
They must wait for a preset number of entries within the database before generating Cpk, rather than selecting a lesser number of current values and quickly determining its impact on Cpk.
(d) Most plant personnel do not realize that they can increase quality to industry standards, yet realize no savings in raw material usage and costs.
(e) Manufacturing facilities typically employ numerous types of coating applications as well as different colors among coatings. Substantial engineering and labor time is allocated for each manufacturing facility to analyze SPC charts that originate from the following categories:
1. each surface area measured;
2. each painting booth within the plant;
3. each style of manufactured product that is produced;
4. each color group or individual color that is used on the painted product; and
5. specific time frames where process improvements or evaluations are being conducted.
(f) substantial coating waste occurs because non-random trends are only identified a substantial period after the coating is applied, thus requiring excessive coating reworks of the painted products.
Typically, a coatings engineer will review an SPC chart, but fail to note the impact that continuing what is noted only as a temporary optimum trend, will have on the process in terms of savings in material usage and costs.
Appendix B highlights this point. This is an actual example from a coating process. Factors causing optimum Range value numbers: 7 (10/31), 8 (10/31), 12 (11/20), 13 (11/23), 14(11/27) and 20 (12/21) are observed but not scrutinized. The respective range values are: 0.01, 0.02, 0.03, 0.04, 0.01 and 0.03. Current commercially available statistical software packages do not correlate the effect that continuing an optimum range will have on reducing coating material usage and costs. At this one surface area, over a two month period (October 10 through December 21), the process was able to produce six optimum ranges which average 0.023 mil. versus an overall range average of 0.21 mil. for the same time period. Range values of coating thickness differ from unit to unit close to a magnitude of ten.
Standard prior art operating procedures continue tracking the data until a shift in the process data warrants the calculation of new control limits. Calculating new control limits usually require justifying to management why new control limits are necessary. Factors that warrant the calculation of new control limits, that are normally considered using the prior art method, include:
1. Waiting for a trend of seven consecutive points moving in the same direction, either upward or downward which indicates a gradual change in the process.
2. Waiting for seven points above the central line which indicates that the center of the normal distribution has started upward.
3. Waiting for seven points below the central line which indicates that the center of the normal distribution has started downward. Source: AIAG (Automotive Industry Action Group) Statistical Process Control (SPC) Reference Manual: pg. 41.
However, what the coatings industry and commercially available SPC programs fail to take into account is the analysis of obvious nonrandom patterns that exist in the Range average charts, specifically, the repeatability and/or pattern of a minimum of two optimum ranges that exist within a sample size of twenty readings. Twenty readings are the minimum number of points required for analysis. (Source: Implementing Six Sigma, Smarter Solutions Using Statistical Methods, by Forrest W. Breyfojlee III, pg. 160)
The AIAG recommends analyzing obvious nonrandom patterns on Range Charts. They recognize the importance of analyzing even a single point, but only if it falls out of the control limits. (Source: AIAG Reference Manual, pg. 42, 43 and 45)
The AIAG and the Coatings Industry have not recognized, the material, cost and environmental benefits of analyzing two optimum ranges found within a sample size of twenty readings, and within control limits.
A software program that analyzes material and/or cost impact based on optimum process ranges would enable engineers to respond more quickly to a detectable trend change. The response would include inserting the optimum achievable target range values into the program and comparing new control limits against industry and/or world class Cpk standards. This allows them to justify to their management, the reasons for identifying the sources contributing to the six optimum ranges, identified in Appendix B. Data can then be used to control the process rather than only tracking the process.
Appendix D, highlights what such an opportunity would provide.
The following data is available from the SPC chart [Appendix B]:
1) Coating
2) Surface Area Measured
3) Booth
4) Coating Minimum Specification
5) Actual Average
6) Actual Range
7) Actual Cpk 
The following data is available from plant operations:
1) Usage Per Unit
2) Coating Cost per Gallon
3) Coating Popularity
4) Measured Surface Area Percentage
5) Booth Flow Percentage
6) Annualized Painted Part Production Volume
7) Industry or World Class Cpk Standard (Available from Reference Manuals)
The engineer inserts into the program data information from an ongoing process pertaining to the optimum process range of film thickness under a 1st Premise, identified as Target Range. New coating average thickness, upper and lower control limits or thickness are calculated as well as the effect of the change on coating usage (gallons) and costs.
The engineer then inserts into the software program information from an ongoing process pertaining to the optimum range pertaining to the coatings applicators manufacturer guidelines under a 2nd Premise, (below 1st premise), identified as Target Range.
The term xe2x80x9cpremisexe2x80x9d means selecting either a new target range, a new range reduction value or a new adjusted coating average.
New coating average, upper and lower control limits are calculated as well as the effect of the change on coating usage and costs.
The novel software program automatically calculates under a 3rd Premise, the effect of maintaining the coating average constant but adjusting variability to the Industry or World Class Standard.
New upper and lower control limits are calculated together with their effect on coating usage (gallons) and costs.
Note this is an important premise. Since in each calculation, if the coating average remains constant, no reductions are obtained in material costs.
The novel process then automatically calculates under a 4th Premise, the effect of adjusting the coating average to the Industry or World Class Standard, but maintaining variability constant.
For example, data reveals that with the first premise for one surface area, a 33.35% reduction in coating usage or $1,418.85 in cost savings could be obtained if the engineer replicates the conditions contributing to the optimum range. Note that this surface area represents 1.05% of one coating. Potentially, if this optimum range existed across the entire painted part, a cost savings of $135,059.05 would be realized. If the cost per gallon is $30.00, 4,502 gallons of coating would be saved. (Appendix D)
Another example of such waste is an industrial facility not realizing that several Coatings"" Film Build Cpks on certain parts of a painted surface have surpassed the Industry Standard. By not identifying the material and financial impact of this achievement, adjustments are not made to the other painted surface areas. The result is the continued acceptance of excess paint usage on this part, resulting in an annual additional expenditure of $764,654.00 instead of product savings of $49,232.00. Assuming a $28.00 cost per gallon, the facility consumed an unnecessary 27,309 gallons of paint using the current method of tracking process data. (Appendix G)
The known prior art fails to address the aforementioned problems. U.S. Pat. No. 5,737,227 to Greenfield et al. describes a software-planning program for coatings but does not correlate the selection of an optimum range found within the process to Cpkxe2x80x2s. U.S. Pat. No. 6,067,509 to Gaiski describes a SPC software program from Pelt Gage Thickness Measurements but does not correlate the selection of an optimum range found within the process to Cpkxe2x80x2s impact on material usage and costs.
The objectives of the present invention are:
to provide an analytical tool for coating and painting facilities:
to reduce variability and improve quality with Cpk Industry Standards and/or World Class Standards;
to generate the usage impact of materials based on the correlation between optimum film build range averages with Cpk Industry Standards and/or World Class Standards;
to generate the cost impact of materials based on the correlation between optimum film build range averages with Cpk Industry Standards and/or World Class Standards;
to generate the usage impact of materials by correlating the process variability remaining constant and adjusting the film build average to Cpk Industry Standards and/or World Class Standards;
to generate the cost impact of materials by correlating the optimum film build range averages obtained from the manufacturing process with Cpk Industry Standards and/or World Class Standards;
to generate the cost impact of materials by correlating the optimum film build range averages obtained from the coatings applicators"" manufacturer guidelines with Cpk Industry Standards and/or World Class Standards;
to generate the usage impact of materials by correlating the optimum film build range averages obtained from the manufacturing process with Cpk Industry Standards and/or World Class Standards;
to generate the usage impact of materials by correlating the optimum film build range averages obtained from the coatings applicators"" manufacturer guidelines with Cpk Industry Standards and/or World Class Standards;
to generate the cost impact of materials by correlating the process variability remaining constant and the film build average adjusted to Cpk Industry Standards and/or World Class Standards;
to minimize the amount of reworks by reducing variability in coating applications by correlating optimum film build range averages with Cpk Industry Standards and/or World Class Standards;
to minimize the amount of paint sludge generated in coating applications by correlating optimum film build range averages with Cpk Industry Standards and/or World Class Standards;
to minimize the amount of volatile organic compounds generated in coating applications by correlating optimum film build range averages with Cpk Industry Standards and/or World Class Standards; and
to minimize the amount of labor required for cleaning paint overspray generated in coating applications by correlating optimum film build range averages with Cpk Industry Standards and/or World Class Standards.
The novel program is used in paint departments to optimize paint usage, to reduce material costs, and to improve the quality of the painted part through variability reduction. Additionally, environmental benefits are achieved by reducing volatile organic compounds, which are directly related to paint usage. Landfill reduction is also achieved as less paint is consumed to generate paint sludge requiring disposal. Plant labor costs for cleaning paint overspray are also reduced.
This new statistical analysis links together a paint department""s quality, financial and process analysis to yield both economic and environmental benefits. The analysis enables coating manufacturing applicators a way to accurately predict the costs and material savings associated with their equipment, using the coating range available with their equipment.
Accordingly, the objects and advantages of the present invention are to reduce material usage and costs by incorporating within a manufacturing facility, a means for generating Film Build Cpkxe2x80x2s Material and or Cost Impact Analysis sheets:
(a) from each surface area measured, thereby reducing material usage and costs;
(b) from each painting location measured;
(c) from each style of manufacturer product measured;
(d) from each color group or individual color measured;
(e) from each coating type measured;
(f) from specific time frames measured;
(g) on a timely measurement basis; and
(h) for each manufacturer of coating applicators based on their particular performance ranges.
Using the preferred method, the Engineer reviews preferably 20 points on an X-Bar R chart. Each point may be the average of all the measurements of sample cars for a day.
The Engineer then selects two points with the lowest ranges from the 20 points. The values of the selected two points are then used to recalculate Cpk, Cp etc.
This is a repeated process, that is, the Engineer reviews each new set of 20 points, takes the two points with the lowest range and repeats his calculation.
In some cases, it may be viable to recalculate Cpk after each new set of a lesser number of points, as long as he considers the latest 20 points. He thus is continuously monitoring the process, and tightening up the tolerance levels by reducing the impact of obvious non-random causes of variability.
Still further objects and advantages of the invention will become readily apparent to those skilled in the art to which the invention pertains upon reference to the following detailed description.